# Can anyone point me to pictures of a region of the night sky during a total eclipse and with the sun not present?

Can anyone show me where I can find at least one photograph of the region of the night sky during a total eclipse and another at night with the Sun not present? I'm interested in the data that astronomers worked with in proving that relativity theory was correct and the challenges they faced. Either the photographs that actually proved the theory or more modern versions would do, preferably brightness/luminosity normalized for comparison, if necessary.

Since science is a complicated business, 'perfect' simulated images of deflected stars versus non-deflected stars using a virtual (i.e. black) Sun (or even a black hole with equivalent mass) in that (or any) portion of the night sky would do for my needs. I'm interested in looking at a composite image with (for example) red pane deflected, blue pane non-deflected ('color differencing') to see if the relativistic deflection would be apparent visually. Thanks.

• Commented Apr 26 at 6:45
• Great link, PM. Includes an interesting discussion that even newbies like me can understand w good analogies. Here's Brun's pdf. arxiv.org/pdf/1802.00343 Since rigorous science is a complicated business, I'll add to my question that 'perfect' simulated images of deflected stars versus non-deflected stars using a virtual (i.e. black) Sun in that portion of the night sky would do for my needs. I'm interested in looking at a composite image with (for example) red pane deflected, blue pane non-deflected (color differencing) to see if the deflection is apparent visually. Commented Apr 26 at 13:23
• The deflection is difficult to see in an image containing the whole Sun. The maximum deflection angle, for a ray grazing the photosphere, is ~1.7512 arcsecs. The deflection is inversely proportional to the distance from the centre of the Sun, so a ray at 1.7512 solar radii gets deflected by 1 arcsec, at double that distance we get half that deflection. But the Sun's angular radius is ~960 ± 16 arcsecs. So you need a lot of pixels to see a deflection of >1 pixel, even for stars very close to the Sun. Commented Apr 26 at 14:25
• A typical eclipse photo won't do as most stars close enough to exhibit deflection would be drown out by the corona at normal magnifications. It really takes a deliberate effort to put the deflected star and reference objects in the field of view at high enough magnification to resolve them precisely. Commented Apr 26 at 16:17
• It's not what you want because the actual deflections are too small to be seen but it is essential that you read and enjoy the classic paper that announced the measurement of deflections. It is (compared to a lot newer and less well written papers) extremely readable and accessible. royalsocietypublishing.org/doi/pdf/10.1098/rsta.1920.0009 It details exactly how they measured the deflection, and ends with the graph with the lines for Einstein, the line for Newton, and the data: clearly matching Einstein. Commented Apr 26 at 20:09

This isn't quite what you want, but perhaps shows what a remarkable observation this was.

Here you can see a simulated solar disc, and a star "d" (without deflection) that is just "grazing" the limb of the sun, and where you would see it, at "e". There is a second star at "f" and it's deflected image at "g". Unfortunately

That's rather hard to see, so let's zoom in on d

I've drawn the stars as circles because the atmosphere and the optics of telescopes mean that they don't appear as points at this magnification. The sun, here, is so magnified that it appears straight. You couldn't actually photograph star d,e because it would be lost in the solar corona.

It would not be possible to see the sun in an image that showed f and g as separate. Here is the zoomed in version.

Of course you can't smootly magnify so your actual before and after images are closer to:

Your challenge is to measure the deflection of the red circle from the blue one.

As you see from these simulations, measuring the deflections requires very careful astrometry, measuring the centre of "fuzzy dot" on a photographic plate or (nowadays) getting a computer to determine the centroid of pixelated image of a star.